[Bug tree-optimization/91071] [10 Regression] Miscompare of 410.bwaves since r272644

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91071

--- Comment #5 from Martin Liška <marxin at gcc dot gnu.org> ---
It's really strange. I isolated the problematic __gcov_one_value_profiler_v2:

__gcov_one_value_profiler_v2 (&__gcov3.flux_[36], ubound.56_347);

but hard to guess why it's causing the issue. I also removed "leaf" attribute
from __gcov_one_value_profiler_v2. And direct return in
__gcov_one_value_profiler_v2 should verify that the fn is not corrupting a
memory.